CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Error: Divergence detected in AMG solver (http://www.cfd-online.com/Forums/fluent/81160-error-divergence-detected-amg-solver.html)

siri October 18, 2010 17:16

DPM Error: Divergence detected in AMG solver
 
Hello friends,

I am encountering either of two errors (given below) when running DPM in Fluent for a turbulent, incompressible, rotating internal flow (like flow inside an annular cylinder).

# Divergence detected in AMG solver: w-swirl -> Increasing relaxation sweeps!
# Divergence detected in AMG solver: k -> Increasing relaxation sweeps!

I am using 2D axisymmetric with swirl option here. Before DPM simulation, I first solved for the flow itself using pressure based solver and obtained a converged solution.

The errors came up immediately after solver restarts from that converged solution. I don't get what they meant. Is it that the k & swirl residuals are too small now as flow has converged and by activating dpm, they will increase backwards. Is that not permitted to do?

Can anyone please explain me what the solver is trying to imply here?

siri October 20, 2010 12:01

any ideas, huh !
 
Quote:

Originally Posted by siri (Post 279690)
Hello friends,

I am encountering either of two errors (given below) when running DPM in Fluent for a turbulent, incompressible, rotating internal flow (like flow inside an annular cylinder).

# Divergence detected in AMG solver: w-swirl -> Increasing relaxation sweeps!
# Divergence detected in AMG solver: k -> Increasing relaxation sweeps!

I am using 2D axisymmetric with swirl option here. Before DPM simulation, I first solved for the flow itself using pressure based solver and obtained a converged solution.

The errors came up immediately after solver restarts from that converged solution. I don't get what they meant. Is it that the k & swirl residuals are too small now as flow has converged and by activating dpm, they will increase backwards. Is that not permitted to do?

Can anyone please explain me what the solver is trying to imply here?

Can anyone respond to this please. guys.... where are you

Laurence Wallian October 21, 2010 05:50

perhaps you can try to reduce some under-relaxation factors...

siri October 21, 2010 14:34

thanks laurence, for stopping by. i did try that ... it doesn't seem to matter but.


All times are GMT -4. The time now is 18:41.